PHYSICS: K.-L. YEN 
93 
It may be seen from the above table that the mean values of R, the ratio 
of the negative to the positive mobilities, are different for the two frequencies 
employed, and it appears as though the mobilities did vary — since the ratio 
could not vary unless either or both of the mobilities did. But the following 
considerations will show that this difference is ascribable to experimental 
variations. In the first place, the results of different experimenters, and 
even those of the same experimenter, show a maximum variation of almost 
30%, for instance, from R = 1.16 to 1.37 (Wellisch), or from R = 1.15 to 
1.93 (Loeb). Both of these authors attribute the variations to external varia- 
tions of their experiments. Thus a slight external variation of some kind 
is liable to cause such a variation in the ratio without the mobilities them- 
selves being varied at all. Furthermore, if the mobilities tend at all to vary 
with field-strength, their variations would be much greater than are mani- 
fested since the field-strength is 168 volt/cm. in one case and 14,160 volt/cm. 
TABLE 2 
Results Obtained from the Measurements on Hydrogen, May- June, 1917 
u+ 
u- 
x+ 
X- 
P 
K+ 
K- 
R 
14,758 cycles, 4000 volts 
5.51 
8.20 
6669 
5668 
748 
5.43 
8.10 
1.49 
5.92 
8.20 
6669 
5668 
748 
5.81 
8.10 
1.38 
5.51 
8.20 
6669 
5668 
746 
5.40 
8.10 
1.49 
8.20 
12.21 
5668 
4723 
518 
5.58 
8.35 
1.49 
14.94 
20.99 
4192 
3524 
290 
5.70 
8.15 
1.41 
14.94 
20.99 
4192 
3524 
300 
5.84 
8.35 
1.41 
Mean 
5.56 
8.19 
1.45 
in the other. Or, it may be said that even if the ratio* does vary about 20% 
when the field-strength is increased from 169 to 14,160 volt/cm. (about 
8333%), it may be considered constant for all practical purposes/ However, 
there were enough of the experimental uncertainties to account for the 
variation. 
Thus the conclusion is that between the field- strength of 168 and 14,160 
volt/cm. the mobility of the positive ions remains absolutely normal, and the 
mobility of the negative ions remains normal also between 134 and 12,870 
volt/cm. 
These results more than amply substantiate those obtained by Loeb and 
it is therefore quite safe to conclude that the evidences obtained so far point 
decidedly in the direction of the small-ion theory. 
b. Hydrogen. The results of the measurements in hydrogen with the high 
frequency high potential field are shown in table 2. 
It may be seen from the table that with a potential gradient of 6669 
volt/cm. or X/p = 14.45, the positive mobility remains normal. The nega- 
